EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Jay112 on April 01, 2016, 09:24:47 pm
-
I have an XBee module that has been working perfectly for months on a custom PCB I created. Recently the XBee stopped working, and I discovered there's a short between the power and ground pins.
Could this be caused by slightly overloading any of the pins, causing a slow deterioration? Or would overloading the GPIO pins not cause a short between power and ground? The only things connected to the unit are the 3.3V power supply (I tested the output voltage both before and after the failure, and it was always at 3.3V), 3 LEDs with 200 Ohms resistors on them, and a buzzer connected via a 2N2222 transmitter. There's a 1K Ohms resistor between the XBee output pin and the transmitter's base pin. (Note: Even in the official XBee tutorials for this device they're running LEDs off the board without any resistors at all).
Although the failure of the XBee caused me to ask this question, I'm curious about the general answer concerning any device/MCU/Arduino/etc.
Thanks!
-
Maybe the 3.3V supply is intermittent (spiked high) or unstable (oscillated) ?
-
Maybe the 3.3V supply is intermittent (spiked high) or unstable (oscillated) ?
Unfortunately I don't have an oscilloscope yet, so I can't check things like that.
Is it common for shorts in the chip to be caused by too much voltage? How high do you think the voltage would have to be to cause something like this? The datasheet says to give anywhere between 2.7-3.6 volts.
-
ESD is very capable of creating shorts, as is HV. Junctions can get punctured,
welded, you name it, all kinds of stuff can happen.
One good way of protecting ESD sensitive pins is using zener arrays. Zeners, because
they operate in avalanche mode, have very fast response times. Fairchild, On Semi
make these arrays.
Plain old HV can be protected a number of ways, clamps, fuse blowers (SCRs), a
number of ways.
Google this "up i/o pin protection esd", lost of info available.
Regards, Dana.
-
ESD is very capable of creating shorts, as is HV. Junctions can get punctured,
welded, you name it, all kinds of stuff can happen.
One good way of protecting ESD sensitive pins is using zener arrays. Zeners, because
they operate in avalanche mode, have very fast response times. Fairchild, On Semi
make these arrays.
Plain old HV can be protected a number of ways, clamps, fuse blowers (SCRs), a
number of ways.
Google this "up i/o pin protection esd", lost of info available.
Regards, Dana.
Thanks for the great info! I'll definitely read that document. Since these XBee units cost $40, I'll probably try implementing some of those protection mechanisms.
-
I discovered there's a short between the power and ground pins.
Did you isolate where the short is coming from? That's the first step to figure out what failed. Power and ground kinda run around and connect to a lot of things. If you think it's a MCU or other IC that is shorted, you should remove it from the board and check it out of circuit. There are a lot of other components that can potentially fail in a short condition. Underrated ceramic caps, for instance. There are probably several caps on your board between power and ground.
-
Perhaps not likely in this case, but "tin whiskers" from lead-free solder could also potentially cause shorts after months or years of error-free operation:
https://www.maximintegrated.com/en/app-notes/index.mvp/id/5250 (https://www.maximintegrated.com/en/app-notes/index.mvp/id/5250)
-
I discovered there's a short between the power and ground pins.
Did you isolate where the short is coming from? That's the first step to figure out what failed. Power and ground kinda run around and connect to a lot of things. If you think it's a MCU or other IC that is shorted, you should remove it from the board and check it out of circuit. There are a lot of other components that can potentially fail in a short condition. Underrated ceramic caps, for instance. There are probably several caps on your board between power and ground.
Everything on my board is still working, and I have another XBee plugged into it and it's working fine. I could test the broken XBee module more, but I don't know how much I would gain from that, plus I'm trying to keep it in good condition because I'm going through the return process now.
I'm not even certain that the XBee was burnt out while it was plugged in and powered on. I'm pretty sure that it was working fine one day, and then I unplugged it in order to test another unit, and then when I went to plug the original unit back in that's when I discovered that strange things were happening. After studying a little about some links posted above, I wonder if maybe it was burnt out from ESD.
What do you guys think about anti-static wrist bands? Do you ever use them, or anything similar?
-
Perhaps not likely in this case, but "tin whiskers" from lead-free solder could also potentially cause shorts after months or years of error-free operation:
https://www.maximintegrated.com/en/app-notes/index.mvp/id/5250 (https://www.maximintegrated.com/en/app-notes/index.mvp/id/5250)
That's an interesting article. I had never heard of tin whiskers before. I only use lead solder myself.
Do you guys ever coat your boards with a protective coating? If so, which one do you think is the easiest for an amateur to use?
-
ESD to an I/O pin while power is applied commonly causes this sort of damage via a process known as latchup (https://en.wikipedia.org/wiki/Latch-up). Its also possible for ESD to blow through the gate oxide or damage a PN junction in such a way that the pin driver or buffer effectively shorts the supply when power is applied, either immediately, or when the logic state is changed.
Its possible to avoid ESD damage without a wristband (or other ESD discharge path for your body) and without an ESD mat on your bench, but it requires the self-discipline to *ALWAYS* anticipate possible ESD hazards and equalise potentials safely. One mistake and you have a dead or damaged device. (e.g. simply passing a board or part to someone else can kill it.) As such, and because I expect occasional mistakes, I use a wrist strap and other appropriate ESD precautions whenever I handle notably ESD sensitive parts or boards.
-
ESD to an I/O pin while power is applied commonly causes this sort of damage via a process known as latchup (https://en.wikipedia.org/wiki/Latch-up). Its also possible for ESD to blow through the gate oxide or damage a PN junction in such a way that the pin driver or buffer effectively shorts the supply when power is applied, either immediately, or when the logic state is changed.
Its possible to avoid ESD damage without a wristband (or other ESD discharge path for your body) and without an ESD mat on your bench, but it requires the self-discipline to *ALWAYS* anticipate possible ESD hazards and equalise potentials safely. One mistake and you have a dead or damaged device. (e.g. simply passing a board or part to someone else can kill it.) As such, and because I expect occasional mistakes, I use a wrist strap and other appropriate ESD precautions whenever I handle notably ESD sensitive parts or boards.
Thanks for the info!
What kind of devices/components do you consider especially susceptible to ESD damage? Is a lot of the concern related to price (like being more careful with a $100 piece than with a $3 piece)?
-
I'm studying some info linked to above about protecting circuits. I was reading about avalanche suppressors, and then read that zener diodes are similar.
I happen to have some 1N5339 5W 5.6V zener diodes. Would it be possible to use that as HV protection? Is the 5.6V breakdown voltage too high for a 3.3V device (though I just read that the XBee Pros are tolerant to 5V)?
Sorry for the newbie questions. :)
-
Ahh... I assumed that it was in continuous use and failed during operation.
Is it possible that you installed the Xbee the wrong way around when powered ?
-
Ahh... I assumed that it was in continuous use and failed during operation.
Is it possible that you installed the Xbee the wrong way around when powered ?
Not this time, but I've done that before, and the units seemed to have survived. :) I've also accidentally plugged them in with 2 pins either too far forward or too far backwards, and they seem to survive that too. Recently I've gotten better with visually noticing when they aren't seated correctly, so that hasn't happened as often.
-
It might not be relevant here but a lot of failed RF gear, is a power transistor collector to emitter short, usually appears as a power rail to earth short (or close to it).